U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021) 335–346
© 2021 WIT P ess, www.wi p ess.com
ISSN: 2056-3272 (pape o ma ), ISSN: 2056-3280 (online), h p://www.wi p ess.com/jou nals
DOI: 10.2495/EQ-V6-N4-335-346
ZERO-ENERGY BUILDINGS IN CITIES WITH DIFFERENT
CLIMATES AND URBAN DENSITIES: ENERGY DEMAND,
RENEWABLE ENERGY HARVEST ON-SITE AND OFF-SITE
AND TOTAL LAND USE FOR DIFFERENT RENEWABLE
TECHNOLOGIES
UDO DIETRICH
REAP Resea ch G oup (Resou ce E iciency in A chi ec u e and Planning),
Ha enCi y Uni e si y Hambu g, Ge many.
ABSTRACT
Ze o-ene gy buildings (ZEBs) ha e no ossil ene gy consump ion; his is achie ed by op imizing he
building and balancing he emaining ene gy needs by enewables. I his ene gy can be ha es ed on-
si e, on he building’s en elope and i s es a e, a ne -ZEB is eached. I supplemen a y enewable ene gy
has o be p oduced o -si e on compensa ing land, he ZEB can be eached wi h such compensa ing mea-
su es (ZEB_CM). Clima e and u ban densi y de e mine how a a ZEB is possible. Tempe a u es ou
o com o ange, lack o dayligh and o e hea ing by sola adia ion may cause ene gy demand while
high insola ion o wind speed deli e s good p econdi ions o p oduce enewable ene gy on less land. A
high u ban densi y a oids u ban sp awl and sa es land ou side o he ci ies ha can be used o o he
pu poses (ag icul u e and ene gy p oduc ion, among o he s). Bu , a a ce ain densi y, ne -ZEB canno
be ealized u he mo e, and compensa ing land is necessa y. The pape in es iga es hese e ec s o
15 selec ed ci ies a ound he globe, co e ing all main clima ic condi ions. Based on design ules ou o
li e a u e and own expe iences, a p o o ypical op imized building is de i ed o each loca ion, and i s
ene gy demand is simula ed. S anda d assump ions o he e iciency o enewable ene gy sys ems a e
used o de e mine he need o land o co e i . Fo di e en u ban densi ies, i can be concluded how a
ne -ZEB is possible; i necessa y, he need o compensa ing land is calcula ed. The esul s show ha
o ci ies wi h mode a e clima es, he o al land use (ci y plus compensa ing land) can dec ease wi h
inc easing u ban densi y i he echnology used o -si e has high e iciency (like PV). On he o he hand,
he o al land use may inc ease ema kably wi h inc easing u ban densi y i he used echnology o -si e
has a low e iciency (like he wind o elec ici y and especially wood pelle s o hea ing). The inal
unde s anding is ha ci ies should mee he ene gy needs on-si e by op imized buildings and s uc u es
plus enewable ene gy p oduc ion (PV on he building’s oo s, geo he mal sys ems, e c.).
Keywo ds: compensa ing measu es, di e en clima es, op imized buildings, u ban densi y, ze o-ene gy
building.
1 INTRODUCTION
Buildings accoun o a conside able sha e o ene gy o hei ope a ion and main enance
leading o a signi ican impac on he en i onmen . I is es ima ed ha 30% o he global
sha e o ene gy is consumed by comme cial and esiden ial buildings leading o 28% o
global emissions (excluding cons uc ion indus y) [1]. The demand o ene gy by buildings
is expec ed o show an upwa d end in he coming yea s. Residen ial and comme cial build-
ings consume app oxima ely 60% o he wo ld’s elec ici y [2]. Eigh y- wo pe cen o inal
ene gy consump ion in buildings was supplied by ossil uels in 2015 (including p ima y
ene gy inpu o powe gene a ion; adi ional use o biomass excluded) [3].
The Pa is Ag eemen on clima e change in he yea 2015 cha ed a new cou se in an e o
o check global wa ming. Sus ainable measu es owa ds ene gy-e icien and low-ca bon
solu ions o buildings and cons uc ion can help achie e he cen al aim o his ag eemen , a
ca bon- ee socie y in a ew decades.
336 U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021)
The concep o ze o-ene gy buildings (ZEBs) is equi ed o educe ene gy consump ion
and b ing down CO2 emission. The ‘Ze o’ e e s o he p ima y (= ossil) ene gy demand o
he se ices ha a e necessa y o gua an ee he use ’s com o inside:
– Hea ing
– Cooling
– Elec ici y o a i icial ligh
– Elec ici y o mechanical en ila ion ( ans)
– Domes ic ho wa e
Tha a ge is achie ed by inc easing he e iciency o he building and balance he ene gy
needs by enewables ha a e p oduced ei he on-si e o o -si e.
Following he de ini ion, i mus be no ed ha use s o a ZEB do no , by a , li e
ca bon-neu al! The elec ici y o he use i sel (‘ enan s’ elec ici y’: PC, se e , gadge s,
TV), he ene gy o anspo a ion, nu i ion, e c. a e NOT included and may no be consid-
e ed negligible.
ZEBs can be di ided in o wo classes: a) ne ZEBs and b) ZEBs wi h compensa ing meas-
u es (ZEB_CM). The di e ence a ises om he loca ion o enewable ene gy p oduc ion. In
case o ne ZEBs, he ene gy demand is co e ed on-si e, wi h sys ems on he building’s en e-
lope and/o on he g ound o he own es a e. In case o a ZEB_CM ha ene gy p oduc ion
on-si e is no su icien o co e he demand, hus supplemen a y compensa ing land/al e na-
i e enewable ene gy sou ces ou side o he own es a e mus be used.
Fo a ne ZEB, i mus be s a ed ha he e is a compe i ion be ween he a ea o usage ha
has an ene gy demand and he size o he building’s en elope and es a e o p oduce ene gy o
co e i . Wi h an inc easing numbe o s o eys, i becomes mo e and mo e di icul o balance
he demand ( he oo a ea o he building is he main a ea o enewable ene gy sys ems, i
does no inc ease wi h he inc easing numbe o s o eys, he es a e emains he same, e c.).
Thus, i can be expec ed ha a ne ZEB is only possible o a ce ain, limi ed numbe o s o eys
and u ban densi y.
In he o he case, when compensa ing measu es a e necessa y, a ZEB_CM will use enew-
able echnologies like PV modules, wind u bines o wood pelle s o ene gy gene a ion on
he compensa ing land.
The ene gy demand o a building is mainly in luenced by i s loca ion and design; he main
aspec s a e clima ic condi ions, u ban densi y, o ien a ion and posi ioning o he building,
numbe o s o eys, cons uc ion mass, window o wall a io, dayligh access, na u al en ila-
ion s a egy, shading sys em, ai igh ness, he mal insula ion, e c.
A second decisi e ques ion is whe he he building’s use s ha e he possibili y o use he
building adap i ely by pe sonally adap able he mos a s, ope able windows, shading sys-
ems, ligh swi ches, e c. and i hey ha e he chance o adap hemsel es o di e en indoo
empe a u es wi h hei clo hing (no d ess code).
This pape is based on he hypo hesis ha he be e he building is clima e and use adap-
i e, he lesse is i s ene gy demand.
2 SCOPE OF WORK/AIM OF THE PAPER
In his s udy, 15 majo ci ies a ound he globe ha ing di e en clima ic cha ac e is ics a e
chosen. Ci ies exhibi ing simila clima ic ea u es a e g ouped o ind simila i ies and di e -
ences. Reykja ik, Oslo and Hambu g a e in colde loca ions, ha ing empe a u es less han 10°C
U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021) 337
o mos o he mon hs and demand hea ing. Chicago and Beijing ha e bo h cold and wa m
mon hs and equi e bo h hea ing and cooling. Cai o and Delhi ha e a ho and d y clima e wi h
he need o cooling. Singapo e, Da es Salaam, Jaka a and San o Domingo a e ho and
humid and need ai -condi ioning almos h oughou he yea . Sydney, San iago, Mexico-Ci y
and Addis Ababa belong o a widely com o able ca ego y.
Fo hese ci ies, he ollowing esea ch ques ions a e in es iga ed:
– Wha is he maximum numbe o s o eys, size o he es a e and u ban densi y o which
i is possible o each ne ZEB?
– Wha is he need o land o compensa ing measu es o di e en u ban densi ies o
ZEB_CM? Wha is he in luence o di e en enewable ene gy echnologies (like PV
e sus wind u bines, e c.)?
3 METHODOLOGY
This esea ch sums up and gene alizes indings based on a uni e si y’s mas e cou se. Di e -
en g oups o in e na ional s uden s wo ked on he di e en ci ies ha ing di e en clima es o
ind he sui abili y o ZEBs in hese ci ies. Bo h ne ZEBs and ZEB_CM we e analysed and
design s a egies we e adop ed.
To acili a e he ask, he s udy is ca ied ou wi h o ice buildings ha a e composed o
s anda d o ice ooms. They ha e s anda d condi ions in use and design ha a e easy o
desc ibe. O cou se, a majo po ion o he buil -up a ea comp ises esiden ial buildings; i
also includes e ails, schools, hospi als and indus ies. Residen ial buildings need less ene gy
han o ice buildings; hus i he ZEB o o ice buildings could be achie ed, hen i could also
be achie ed o esidences.
Fu he mo e, eal u ban si ua ions we e no ega ded since his is oo complica ed. A qua -
e o (iden ical) o ice building is assumed, well e lec ing he e ec o educed dayligh
access and inc eased elec ici y demand o a i icial ligh i buildings a e nea oge he and
shade each o he . I was men ioned ha such mono-use qua e s a e no sus ainable. As he
scope o he wo k is o ob ain gene al endencies and compa able esul s wi h simila assump-
ions o di e en ci ies wi h di e se clima ic condi ions, i is di icul o conside he eal
u ban si ua ion. U ban densi y is exp essed as he plo a io, which is he a io be ween he
a ea o usage and he a ea o es a e.
3.1 S anda d o ice oom size and equipmen
The s anda d o ice oom is a uni o 168 m2 wi h a dep h o 14 m and a wid h o 12 m, sup-
posed o p o ide wo king space o 12 people. This model size could be eplica ed h oughou
he whole wid h and heigh o he buildings (wi h supplemen ing s ai s and ele a o s e c.).
Each s o ey heigh / loo heigh in he building is 3.2 m. The ime o usage o he building is
11 hou s and 5 days a week which is om mo ning 7 am o e ening 6 pm om Monday o
F iday.
3.2 Renewable ene gy p oduc ion
3.2.1 On-si e
The he mal ene gy is ob ained by a geo he mal sys em (see sec ion 3.2.3) ha ills in
maximum he whole es a e. The coe icien s o pe o mance (COP) o cooling and hea ing
338 U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021)
a e 2.5 and 3.5, espec i ely. The size o he sys em de e mines he possible maximal powe
o he hea ing o cooling sys em.
The elec ici y necessa y o hea pumps, a i icial ligh ing and en ila ion is ecei ed by
ha es ing sola ene gy h ough PV panels. Polyc ys alline PV placed la on he oo is used
o he analysis. I is assumed ha a su plus can be deli e ed o he g id and a supplemen a y
need could eed ou o he g id, hus he a ea o he PV modules is de e mining he annual
con ibu ion o he elec ici y demand o he building.
Fo a ne ZEB, he desc ibed enewable ene gy sys ems can deli e he whole ene gy
demand. I is de e mined up o which numbe o s o eys o he building and up o which
building dis ance/u ban densi y ha is possible.
3.2.2 On compensa ing land
In case o he assumed numbe o s o eys o he u ban densi ies being highe han he h esh-
old o a ne ZEB, a lack o hea and/o elec ici y is caused. To balance i , compensa ing land
is needed o accomplish he ene gy demand o he building. Compensa ing elec ici y demand
is ob ained om onsho e o o sho e wind u bines o PV modules and supplied o he o ice
building.
T anspo ing he mal ene gy (hea ) om i s sou ce o gene a ion o he place o in ended
use is no p e e ed as i leads o a lo o ene gy losses du ing i s anspo . Hence, anspo
o aw ma e ial on-si e o bu n he e o hea ene gy is mo e economical. Wood pelle s a e
chosen as a compensa ing measu e o he mal ene gy demand.
The e o e, he a ea o compensa ing land depends on he annual ene gy demand as well as
on he chosen ype o enewable echnology.
3.2.3 The ene gy densi y o di e en enewable ene gy sys ems
Sola adia ion deli e s – depending on he loca ion – a ew hund ed o mo e han 2000
kWh/y ene gy o a squa e me e o ea h’s su ace. Renewable ene gy sys ems ans e a pa
o i in o usable ene gy (elec ici y, hea , ma e ial o bu n). Figu e 1 illus a es ha he e i-
ciency o his p ocess di e s ema kably om sys em o sys em ( alues based on [4]).
Geo he mal sys ems ha e highes e iciency (wi h dis ance) o hea ing and cooling and PV
modules o elec ici y. Wind u bines onsho e o o sho e occupy a g ea e amoun o land.
Bu i mus be no ed ha in some cases, i is di icul o ha e any o he use (like ag icul u e)
apa om PV modules in he land ha is co e ed wi h PV as he land unde PVs is da k and
d y. Ag i-PV is an op ion bu , in any case, he e a e wo sys ems, plan s and PV, in compe i-
ion o sola ene gy. The land in be ween he wind u bines in a wind a m could be used
e ec i ely o o he pu poses.
Ene gy plan s hold he lowes ene gy densi y. Ene gy plan s c ea e compe i ion o ag icul-
u e o ood p oduc ion and i should be a oided. Wood pelle s a e a by-p oduc o enewable
o es y; besides he limi ed p oduc ion, he e is no nega i e impac .
These alues gi e he yea ly ha es o enewable ene gy. The sola o e o 953 kWh/m²
y and he ha es o PV modules (e iciency abou 15%) e e o he loca ion o Hambu g,
Ge many. Fo all o he ci ies unde in es iga ion in his a icle, hese alues a e adap ed o he
co esponding sola o e a he chosen loca ion.
Da a o wind e e o an a e age wind eloci y o 3 m/s on land and 5 m/s on he sea and
a e assumed he same o all loca ions.
Own es ima ions show ha he ha es o geo he mal sys ems a ies only sligh ly wi h he
empe a u e in he g ound and hus he loca ion. The sys em is assumed as 100 m deep e ical
U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021) 339
bo ehole hea exchange s wi h a dis ance o 7 m. Wi h he assump ions ha 1 m o hea
exchange can deli e 600 Wh/d he mal ene gy and ha he sys em is unning 6 mon hs a yea
(hea ing o cooling season), an ene gy densi y o 223 kWh/m² y can be de i ed. This alue is
only calcula ed o ha e a ough imp ession o he po en ial o geo he mal sys ems in compa i-
son o he o he sys ems ha a e included in Fig. 1. O cou se, he eal po en ial o he geo he mal
sys em will be di e en o di e en loca ions. I is calcula ed on a daily basis compa ing he
daily hea ing/cooling demand and he maximal capaci y o he geo he mal sys em.
3.3 The in e na ional s yle oom
In a i s s ep, he s anda d o ice oom was assumed as ealized in common a chi ec u e, as
an ‘in e na ional s yle’ oom:
– N-S o ien a ion
– Fully glazed and sealed acades wi h double hea p o ec ion glazing
– In e nal shading sys em
– Ai -condi ioning (26°C), mechanical en ila ion and a i icial ligh du ing he whole
ime o usage
I s ene gy demand o all he 15 loca ions was simula ed wi h P ime o-Com o [5], a an-
sien simula ion so wa e. Resul s show ha o all loca ions, mo e o less, i is no possible
o each a ne ZEB wi h a sa is ying u ban densi y; i such a densi y is assumed, he esul ing
need o compensa ing land would be immense. Tha deli e ed con ic ion and mo i a ion
ha he oom should be op imized and adap ed o he clima e o educe i s ene gy hunge and
o imp o e he chances o each a ne ZEB o a ZEB_CM wi h sa is ying condi ions.
Figu e 1: Ene gy densi y o di e en enewable ene gy sys ems. Da a o sola o e and
esul ing ha es o PV e e o Hambu g, Ge many. The alue o geo he mal is
jus o show he dimension o compa ison.

340 U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021)
3.4 Op imiza ion o he s anda d o ice oom o a clima e adap i e one
To ind he way o an op imized and adap i e building, se e al sou ces we e used:
– Clima e consul an so wa e [6] is used o unde s and he empe a u es, sun shading,
sky co e a io, wind eloci y and humidi y.
– Design s a egies we e de i ed based on he ules gi en by he cou se supe iso s and
hose om he clima e consul an .
– Ve nacula a chi ec u e and bes p ac ice examples.
– Also, he opinions and sugges ions o he s uden s om hese loca ions we e conside ed.
Fi s ly, based on he clima ic condi ions, i was o be decided which mon h o a yea he
building can un in which o he ollowing modes:
– Adap i e: Indoo com o can be main ained wi h only na u al en ila ion and hea ing.
I is assumed ha a building can be un adap i ely i he mon hly mean alues o
ou doo empe a u e lie be ween 10 and 23°C.
Ci ies like Addis Ababa, Mexico-Ci y, Sydney and San iago ha e a high po en ial o
be un he whole yea adap i e.
– Ai -condi ioned: Indoo com o can be main ained only wi h mechanical en ila ion
and cooling.
I is assumed ha a building can be un only ai -condi ioned i he mon hly mean
alues o he ou doo empe a u e is abo e 23°C o a oid indoo empe a u es ou o
he com o ange.
Singapo e which has a ho and humid clima e wi h an a e age yea ly empe a u e o
27°C uses ai -condi ioning o 12 mon hs o he yea . San o Domingo, Jaka a and Da
Es Salaam also need ai -condi ioning nea ly h oughou he yea .
– Hea ing: Indoo com o can be main ained only wi h hea ing and mechanical en ila-
ion wi h hea eco e y o educe ene gy demand.
I is assumed ha a building can be un only in his mode i he mon hly mean alues
o ou doo empe a u e lie below 10°C.
Reykja ik, Oslo, Chicago, Beijing and pa ly Hambu g ha e such s ong win e pe i-
ods. Bu he e is no loca ion whe e ha mode is necessa y o he whole yea .
– Hyb id: Indoo com o can be main ained seasonal by unning he building adap i ely,
wi h ai -condi ioning o hea ing.
Buildings could adap o he su oundings o a ew mon hs and depend on cooling
o hea ing and mechanical en ila ion o he es . Chicago is he bes example; i can be
adap i e o 6 mon hs and needs mechanical en ila ion/hea ing o he es o 6 mon hs
o he yea . Hambu g and Beijing also lie in he ca ego y o hyb id.
Finally, he a chi ec u e could be adap ed o he chosen modes (Fig. 2):
– Adjus men ( educ ion) o oom dep h o be e na u al c oss en ila ion (i o ad an-
age – loca ions wi h weak wind eloci ies) while e aining he a ea o usage o 168 m²
(an inc ease o oom wid h).
– Reduc ion o he window o wall a io in a way ha o e hea ing p o ec ion and dayligh
access is in good symbiosis. The a io o loca ions wi h a dominan cloudy sky is abou
50%, and o a dominan clea sky, i is 35%.
U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021) 341
– Con enien (ex e nal) shading sys em, glazing and he mal insula ion.
– In elligen size and placemen o (ope able) windows o dayligh and na u al en ila-
ion.
– Na u al en ila ion s a egy, especially o nigh cooling.
– A i icial ligh can be swi ched o i dayligh is su icien (500 lx).
These ooms we e p esen ed a he end o he uni e si y cou se and hen u he de eloped
by he au ho o gua an ee ha hey a e op imized o all loca ions a a co esponding le el.
Finally, he ene gy demand o hese ooms was simula ed wi h he same so wa e.
4 RESULTS
4.1 CO2 educ ion po en ial o op imized buildings
Figu e 3 gi es an o e iew o he possible educ ion po en ial in p ima y ene gy (p ima y
ene gy ac o o hea ing 1.1, o elec ici y 3.0) be ween an in e na ional s yle building and
a building ha is op imized and adap i e. I can be seen ha he po en ial (and hus he educ-
ion o CO2 emission!) is eno mous.
I is also e iden om Fig. 3 ha he educ ion po en ial is di e en o di e en loca ions.
I cooling demand is caused by sola hea gains, i can be educed (o b ough o ze o,
Hambu g, Oslo, Sydney) by a chi ec u al means. I i is caused by empe a u es abo e he
com o ange, a chi ec u al means can ha dly help, and cooling is necessa y (Jaka a, Delhi,
e c.). I he loca ion is a om he equa o , he e a e many hou s o usage whe e i is da k
ou side – he demand o a i icial ligh canno be educed (Reykja ik, Oslo, e c.).
Figu e 2: Examples o in e na ional s yle ( o ind in any ci y) e sus op imized and adap i e
buildings o di e en loca ions and clima es (s uden s wo k).
342 U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021)
To deli e a ai compa ison, i is assumed ha he op imized building has mechanical
en ila ion and is cooled oo (i necessa y). He e is u he op imiza ion po en ial; buildings
can o en be un in adap i e mode (see sec ion 3.4) and wi hou mechanical en ila ion.
4.2 Maximal u ban densi y (plo a io) o ne ZEBs in he chosen ci ies
I can be calcula ed up o which u ban densi y (plo a io), ne ZEB is possible o he di e -
en loca ions. The p e equisi e ha he ene gy demand mus be co e ed by enewable ene gies
gained on si e se s wo limi s:
– The yea ly ha es o a PV module sys em on he building’s oo can co e he elec ic-
i y demand (mechanical en ila ion, a i icial ligh , hea pump) o a ce ain numbe o
s o eys.
– A bo ehole hea exchange sys em in he g ound can deli e a ce ain amoun o powe
o hea ing o cooling. Wi h i , he hea ing and cooling demand o a ce ain numbe o
s o eys can be co e ed. Fo he calcula ion, a building dis ance o 20 m is assumed,
which co esponds o a ypical and ealis ic s ee wid h. F om his assump ion and he
size o he es a e, he size o he geo he mal sys em can be de e mined. In case i is in-
su icien and he e is a ese e in he PV elec ici y p oduc ion, s anda d chille s (COP
= 1.5) a e u he assumed un il he PV sys em is exhaus ed.
The dominan o he wo c i e ia decides abou he maximal numbe o s o eys. The esul -
ing u ban densi y can be inally calcula ed. Figu e 4 shows he esul s.
Figu e 3: Compa ison o he p ima y ene gy demand o in e na ional s yle and op imized
building.
U. Die ich, In . J. o Ene gy P od. & Mgm ., Vol. 6, No. 4 (2021) 343
F om Fig. 4, i is no iceable ha Reykja ik has he lowes plo a io 0.8 which e lec s
un a ou able condi ions p e ailing in he ci y: a s ong win e causing high hea ing demand,
ex ended imes o da kness causing demand o a i icial ligh and oge he wi h he mechan-
ical en ila ion leading o high elec ici y demand – on he o he hand, i has he lowes sola
adia ion and he sun posi ion ha dly abo e he ho izon ha esul s in a e y low ha es wi h
PV modules on he oo .
The highes possible plo a io o 3.4 is shown by Addis Ababa which has a mode a e cli-
ma e. A mode a e clima e equi es nea ly no he mal ene gy o hea ing o cooling, dayligh
in he whole ime o usage, no mechanical en ila ion leading o a minimal elec ici y demand
ha can easily be co e ed wi h he PV sys em on he oo ecei ing a high amoun o sola
adia ion. I is no ed ha he loca ions wi h simila clima es show also simila esul s he e. O ,
ice e sa, he local clima ic condi ions de e mine he chances o each bo h, ne ZEB and a
highe u ban densi y.
4.3 Need o compensa ing land o ZEB_CM o selec ed u ban densi ies (plo a io) o
he chosen ci ies
I he planned u ban densi y exceeds he limi o ne ZEB, supplemen a y compensa ing land
o enewable ene gy p oduc ion is necessa y. Tha land mus be loca ed ou side he ci y (o
on he sea in case o o sho e wind u bines). I mus be e alua ed p ecisely i and whe e ha
is a ailable. Compe i ion be ween ene gy and ood p oduc ion mus be a oided. On he o he
hand, a ci y wi h a high u ban densi y a oids u ban sp awl and sa es land; a compac ci y has
sho dis ances and sa es ene gy o anspo a ion. The bes combina ion o each loca ion
mus be ound by balancing all impac s.
Figu e 4: Maximal u ban densi y (plo a io) o ne ZEB o he chosen ci ies.